Concrete surface defects diagnosis field guide

Why the defect is the diagnosis

A concrete surface defect is evidence. The slab is telling you what went wrong, and if you read it instead of guessing, you fix the process instead of repeating it. Scaling, crazing, dusting, blistering, delamination, popouts, each one points at a different mistake in the mix, the finishing, the cure, or the bleed water. The wrong blame repeats the failure on the next pour. The right diagnosis ends it.

Most surface defects share a short list of roots. Finishing the slab while bleed water still sat on it. No air for the freeze and the deicer. Sealing the top before the concrete underneath was done bleeding. No cure, so the surface dried before it hydrated. The finishing-sequence guide covers the timing that prevents the finish defects, and the curing guide covers holding the moisture that prevents the cure defects. This guide runs the other direction, from the failure on the floor back to which of those went wrong.

Diagnose before you repair. A grinder and an overlay hide almost anything for a season, but patch a scaled slab without fixing the air content or the finishing and the patch scales too. The repair is the cheap part. Knowing why it failed is the part that keeps it from failing again, and it is the part the owner is paying for when they call you back.

The bleed water and the timing behind most defects

Bleed water sits under more surface defects than any other single cause, so start there. After the concrete is placed and struck off, the solids settle and water rises to the top. That bleeding has to finish, and the surface water has to leave, before any floating or troweling touches the slab. Work the surface while the water is still there and your tools push it back down into the top fraction of an inch, where it has no business being.

What you get is a thin, soft, high water-to-cement skin, and any air the mix carried trapped right under a sealed top. That one mistake produces four different failures depending on the mix and the exposure. It dusts under traffic. It scales in freeze-thaw. It blisters as the trapped water and air rise. It delaminates in sheets. Same root, four faces.

Timing is the whole game, and the clock does not call it. A hot, dry, windy day pulls the bleed forward and stacks the steps. A low-bleed mix heavy in fly ash, slag, or silica fume barely bleeds at all, so a crew waiting for a water sheen that never comes holds on a surface drying and cracking underneath. The set test, not the clock, opens the window. The finishing-sequence guide covers reading that surface. Here the point is only that almost every finish defect traces back to working the slab at the wrong moment.

What causes concrete to scale?

Scaling is the flaking or peeling of the top layer of a slab, commonly the upper 1/8 to 3/16 in, leaving the mortar and sometimes the coarse aggregate exposed. It shows up almost entirely on exterior flatwork that sees cyclic freezing and thawing, and it gets worse fast where deicing salts are used. Scaling is the most common cold-climate surface failure, and it is a durability problem, not a cosmetic one.

The leading cause is air. Concrete exposed to freeze-thaw needs an adequate system of entrained air voids so the water in the surface has somewhere to go when it freezes and expands. No air entrainment, or not enough for the exposure, and the freezing water has no relief. The pressure builds and the surface flakes off. Deicers make it worse by driving more freeze-thaw cycles and adding salt pressure on a surface that was already short on air.

Finishing is the other half. Finishing over bleed water leaves a weak, high water-to-cement skin that scales easily, and overworking the surface drives the entrained air out of the top layer exactly where it is needed most. So a slab can be batched with good air and still scale if the finisher closed it wrong. Air-entrained concrete resists deicers and freeze-thaw if it is properly finished and has reached maturity before its first exposure, which is why a new exterior slab should not see deicing salt its first winter. Verify the air content for the exposure class with the mix supplier and against ACI 201 durability guidance, hold the finishing to the bleed-water rule, and cure it.

What is crazing, and does it matter?

Crazing is a network of fine, shallow, random hairline cracks across the surface, in a pattern people compare to a dried lakebed, a road map, or a cracked windshield. The cracks are rarely more than about 1/8 in deep and do not reach into the body of the slab. They are most visible on hard-troweled surfaces, and they jump out when the surface is damp or after a sealer goes on.

The cause is the surface drying and shrinking faster than the concrete under it. Low humidity, wind, and direct sun on a fresh pour pull water off the top, the thin surface layer shrinks against the stable concrete below, and it cracks in the random map pattern. Over-finishing feeds it by bringing extra fines and water to the surface, and so does the old habit of sprinkling water or dry cement on the slab to make it trowel easier. No cure lets the surface dry too fast and seals the result in.

Here is the honest part: crazing is cosmetic. The cracks are too shallow to affect strength or, in most cases, durability, so on a utility slab it is a blemish you note and move past. It matters where appearance does. On a slab to be polished, stained, or left exposed in a finished space, the map pattern reads as a defect and the owner sees it. Treat crazing as a flag that the surface dried too fast or was over-finished, prevent it by curing early and finishing with a lighter hand, and do not tear out a sound slab over it.

Why is my concrete dusting?

Dusting is a soft, powdery, chalky surface that rubs off under traffic or comes away on your hand. It is a thin top layer that never reached real strength, sitting over sound concrete below, and it keeps shedding fine powder until something is done about it.

The most common cause is finishing bleed water into the surface. Working that water back into the top 1/4 in raises the water-to-cement ratio right where wear hits and leaves a weak, low-strength skin. Carbonation is the winter-enclosure cause: carbon dioxide from unvented salamanders, gas heaters, generators, or power buggies reacts with the calcium hydroxide in fresh concrete to form a soft, carbonated, dusting skin. Add a high water-to-cement mix, no cure so the surface dried before it hydrated, condensation troweled into the surface in humid conditions, and a non-absorptive subgrade or vapor retarder that increases bleeding, and the surface comes up weak from any of several directions.

The fix depends on how bad it is. A lightly dusting floor can be hardened with a chemical densifier, a silicate that reacts and toughens the surface. A badly dusting floor gets ground off and overlaid, because there is no sound skin to save. Prevent it the same way every time: do not finish over bleed water, vent every combustion engine and heater in an enclosed pour, use a low enough water content, and cure the slab. The curing guide covers holding that moisture so the top hydrates instead of dusting.

What causes blisters in concrete?

Blisters are small hollow bumps in a troweled surface, from dime size to a few inches across, that rise during or shortly after troweling. Break one open and the shell is dense and the inside is hollow, with a thin layer of mortar lifted off the concrete below. They show up on troweled interior floors, and they are an early warning of the wider failure, delamination.

The cause is sealing the surface too early. The finisher trowels and densifies the top while the concrete underneath is still bleeding and releasing air. The rising bleed water and air cannot get out through the closed skin, so they collect under it and lift the surface into a blister. Anything that crusts the top while the body is still working raises the risk: a cohesive or sticky mix, a cold subgrade that stretches the bleeding while warm air, sun, or wind sets the surface fast, and most reliably, air-entrained concrete given a hard steel-troweled finish. ACI floor and slab guidance recommends against hard-troweling air-entrained concrete for exactly this reason.

Small blisters can sometimes be flattened if they are caught the moment they appear and the surface is reopened with a float and the troweling delayed, but once the slab has set, a blister is a void. The real fix lives in the finishing timing, not in chasing bumps. Wait for the bleed sheen to leave and the surface to take a footprint around 1/4 in deep before power floating, keep a magnesium float on the early passes to leave the surface open, and do not hard-trowel air-entrained concrete. The finishing-sequence guide carries the full timing.

Delamination: the hollow top layer

Delamination is a blister grown wide. A thin densified top layer, often 1/8 to 1/4 in thick, debonds from the concrete under it over a broad area. The surface looks fine and tight at first. Then it sounds hollow when you drag a chain across it, tap it with a hammer, or roll a steel rod over it, and eventually it breaks under a forklift wheel or foot traffic and the whole sheet lifts.

The cause is the same as blistering, spread over a large area: troweling over rising bleed water and air, sealing the surface before the concrete below was done bleeding and releasing air. The trapped water and air collect in a plane right under the closed skin, and that plane has no bond. Delamination is most likely when something extends the bleeding time, like a cold substrate, while something else speeds the surface set, like high air temperature, sun, or wind, because the top crusts while the body is still bleeding. Air-entrained concrete hard-troweled is the classic setup, since the entrained air has nowhere to escape.

Find it by sounding before it shows. Drag a chain or a wheeled sounder across a suspect floor and map the hollow areas on the plan, because the visible breaks are usually smaller than the debonded zone around them. There is no patch that bonds to a hollow you leave in place. The repair is to remove the delaminated layer back to sound concrete and overlay, after a bond test, which is why this defect costs real money. It is a finishing-timing failure, and the prevention is the bleed-water rule and no hard trowel on air-entrained slabs.

Plastic shrinkage cracking and the evaporation rate

Plastic shrinkage cracks are short, random cracks that open in the surface within the first 1 to 6 hours after placement, while the concrete is still plastic and has no real tensile strength. They often run roughly parallel, a foot or two apart, and they do not reach full depth. You see them on hot, dry, windy pours, opening before the slab has set, sometimes while the finishers are still on it.

The cause is rapid evaporation. When water leaves the surface faster than bleed water can rise to replace it, the top layer dries and shrinks while the concrete underneath holds it back, and the surface tears. ACI hot-weather guidance puts the action level at an evaporation rate approaching 0.2 lb per square foot per hour, and recommends precautions in the 0.1 to 0.2 range. That rate is driven by air temperature, concrete temperature, relative humidity, and wind, and a windy low-humidity afternoon reaches it well before the air feels that hot.

Caught early while the concrete is still plastic, the cracks can sometimes be closed by re-floating and re-finishing. Once the slab sets they are permanent, and while they are usually a durability and appearance issue rather than a structural one, they open a path for water into the surface. The prevention is staged before the truck on a hot pour: fogging to raise the humidity over the slab, windbreaks, shade, and an evaporation reducer sprayed on the surface to slow the loss. The curing guide and the plastic-stage protection it tracks cover reading the evaporation rate and holding the surface.

Popouts: the conical holes and the aggregate at the bottom

A popout is a small conical depression in the surface, from a fraction of an inch to a couple of inches across, with a fractured piece of coarse aggregate sitting at the bottom of the hole. The cone shape and the aggregate are the tell. This is not a finishing or curing defect. It is the aggregate itself failing near the surface.

Two mechanisms drive it. The physical one is an unsound, porous, or absorptive aggregate particle near the surface that takes up water and expands when it freezes, fracturing and lifting the mortar above it. Frozen aggregate placed in the mix does the same on a smaller scale. The chemical one is alkali-silica reaction, where reactive silica in certain aggregates reacts with the alkalis in the cement paste to form a gel that swells with moisture and pushes the surface off. Either way the fault is in the stone, not in the crew.

Popouts are mostly cosmetic on a slab, though they are a real problem on a floor that has to be smooth and a sign of a deeper durability issue if alkali-silica reaction is at work. The prevention is in the mix, not the finishing: source sound, non-reactive aggregate, and for an aggregate known to be reactive, use a low-alkali cement or a supplementary cementitious material to suppress the reaction. This is a conversation with the mix supplier and, where reaction is suspected, a petrographic look at the aggregate, not something the finisher can fix on the slab.

Discoloration: blotchy, dark, and mottled

Discoloration is uneven color across a slab that should be uniform: dark blotches, light patches, or a mottled pattern. It is almost always cosmetic, but on architectural and decorative concrete it is the whole job, so it gets treated as a defect that has to be tracked to a cause.

The causes are a list of small inconsistencies. Calcium chloride accelerator darkens concrete, so a load with chloride and a load without will not match. A change in the cement source or the aggregate mid-pour shifts the color. A variable water-to-cement ratio across the slab, from adding water to part of a pour, leaves the wetter areas a different shade. Hard troweling and over-troweling burn the surface dark and blotchy, a near-black marbling that does not come out. And uneven curing prints on the surface: plastic sheet in direct contact mottles the color where it touches versus where an air gap forms, and patchy curing compound coverage leaves its own pattern. The curing guide covers the plastic-mottling and coverage problems in full.

Discoloration is hard to remove once it is in the surface, and the realistic answer is often consistency going forward rather than a cure for the slab in front of you. Keep one cement source and one admixture dose across a pour, hold the water-to-cement ratio steady, do not burn the surface chasing a sheen, and cure the whole slab the same way at the same time. Where color is the deliverable, that consistency is the work.

Curling: why the slab edges lift

Curling is the upward warping of slab edges and corners, most visible along the joints, where the panel lifts off the subgrade. You feel it before you see it, as a bump a wheel hits crossing a joint, and over time the lifted edges chip and the joints spall under traffic.

The cause is differential shrinkage from a moisture gradient. The top of the slab dries to a lower moisture content than the bottom, so the top shrinks more, and the slab warps upward at the edges where it is free to move, which is at the joints. A wet subgrade under dry interior air sets up the gradient, and a high-shrinkage mix, a thin slab, and wide joint spacing all make it worse. It is the same drying shrinkage that opens the random map of cracks elsewhere, only here it lifts the panel.

Curling sits between cosmetic and functional. On a back-of-house slab it is a nuisance. On a floor with traffic across the joints, the lifted edge becomes a maintenance and a wheel-load problem, and it is the defect a high-flatness floor cannot tolerate. Reducing it is a mix and detailing question more than a finishing one: a lower-shrinkage mix, sensible joint spacing, and attention to the vapor and subgrade conditions. The curing guide and the control-joint layout both bear on it, because how the slab dries and where the joints fall decide how much it curls and where it shows.

Is the defect cosmetic or structural?

Sort the defect into one of three buckets before you decide anything else, because the bucket sets the response. Cosmetic defects are surface appearance only: crazing, light dusting, popouts, and discoloration look bad and change nothing about whether the slab carries its load or lasts. Surface-durability defects are the wear surface failing: scaling, blistering, and delamination mean the top is coming apart, so the floor will not last under traffic or weather even though the slab below is sound. Structural defects are a different family, full-depth cracks, settlement, and through-cracking that touch the capacity of the slab, and they are the engineer's call, not the finisher's.

The diagnostic moves are simple and physical. Check depth: a shallow surface pattern is cosmetic, a crack you can trace down into the slab is not. Sound it: a hollow ring under a chain or a hammer means delamination, a solid ring means the surface is bonded. Read the pattern: a random shallow map is crazing, a conical hole with aggregate is a popout, parallel early cracks are plastic shrinkage. Ask what the defect touches, the look of the surface, the life of the wear surface, or the strength of the slab.

When it reads structural, stop diagnosing and bring in the engineer. A finisher can call scaling, dusting, and crazing all day, and should. A through-crack across a structural slab, signs of settlement, or cracking near columns and loads is not a surface defect, and guessing at it is how a cosmetic conversation becomes a liability. Know where your call ends.

Defects on flat floors and data center slabs

On an ordinary slab a little curling or a small delaminated patch is a minor problem. On a superflat warehouse floor, a data center slab, or any high-flatness floor, the same defect becomes functional, because the tolerance is tight and the equipment is unforgiving. A high-speed turret truck in a very narrow aisle feels a curled joint as a jolt that throws the mast. A precision rack or a row of cabinets sits wrong over a waved or curled surface.

Flatness and levelness are measured as FF and FL numbers under ASTM E1155, and the defect that hurts a flat floor most is curling at the joints, which shows up as flatness lost after the slab dries even when the floor measured flat green. That is why a flat floor is read within about 72 hours of placement, before curling drifts the numbers, and why the curling defect gets weighed against a much tighter spec than ordinary flatwork. A delamination on a data center floor is not just a wear problem, it is a flatness and a contamination problem under the raised floor or the racks.

The flatness itself is built in the screed and the floating, not the trowel, so a flat-floor defect usually traces to the placement and the restraightening rather than to the final passes. The finishing-sequence guide covers building the FF and FL numbers. The diagnostic point for this guide is that the same defect carries a different weight depending on the floor: judge it against what the floor has to do, not against a generic slab.

Prevention: the common thread

One thread runs through almost every defect in this guide, and it is short. Get the mix right for the exposure, wait for the bleed water to leave, do not seal the surface early, keep finish water off the slab, and cure it. Do those five and the dusting, scaling, blistering, delamination, and most of the crazing never start.

The mix carries the defects you cannot fix at the trowel. Air entrainment for any concrete that will see freeze-thaw and deicers, or it scales. A low enough water-to-cement ratio, or the surface comes up soft. Sound, non-reactive aggregate, or it pops out. Those are decided with the mix supplier and against the durability guidance in ACI 201 before the truck is ordered, and no amount of good finishing recovers a mix that was wrong for the exposure.

The finishing and the cure carry the rest. Bull float before bleed, wait out the bleed, finish to the schedule, and never work water or dry cement into the surface to make it trowel, all covered in the finishing-sequence guide. Then cure the moment the finish is done, holding the moisture so the surface hydrates instead of dusting, scaling, or crazing, which the curing guide covers in full. Blaming the mix for a finishing defect, or the finisher for a mix defect, is how a crew fixes the wrong thing and pours the same failure again.

Repair or replace the slab?

Match the repair to the bucket, and never repair before you have found and fixed the cause. Cosmetic defects, crazing, discoloration, and light popouts, usually get accepted, sealed, or stained over, because there is nothing failing to fix. A lightly dusting surface gets a chemical hardener. Those are touch-ups, and they hold because the concrete under them is sound.

Surface-durability defects need the failed layer gone. A scaled, badly dusting, or delaminated surface gets ground or shot-blasted back to sound concrete and overlaid or resurfaced, after a bond test and proper surface profile, because an overlay bonded to a weak or hollow skin debonds right along with it. This is the most common repair mistake in the trade: skim-coating over a delamination you never removed, so the new surface lifts with the old one inside a year. Sound the floor first, remove every hollow area, prep the substrate, then bond to it.

Structural defects are not a finisher's repair at all. Full-depth cracks, settlement, and anything touching the load path go to the engineer, who decides between a structural repair and replacement. The honest framing for an owner is that the slab almost never has to be replaced for a surface defect, only resurfaced, while a structural problem is a different and bigger decision. Either way, fix the process that caused it before you spend a dollar on the repair, or you are buying the same callback twice.

Mapping and photographing the defect

Document the defect before you touch it, because the repair erases the evidence and the cause argument happens later. Photograph it with something for scale in the frame and enough context to place it on the floor, and shoot the surface both dry and wet, since crazing and discoloration read differently when damp. A photo with no scale and no location is a photo nobody can use in six months.

Map the extent, especially for delamination. Sound the floor with a chain or a rod, mark the hollow areas on a floor plan, and record the total area, because the visible break is smaller than the debonded zone and the repair has to cover the whole hollow. For scaling and crazing, outline the affected panels on the plan so you can tie the defect to a pour, a truck, or a stretch of weather.

Then write the probable cause while it is in front of you. The finishing-sequence record of when the bleed left and when finishing started, and the curing record of method and start time, are what let you confirm whether this was a finish defect or a cure defect rather than guess. A defect captured with a photo, a map, and a cause note is a defect you can diagnose, defend, and prevent. One captured from memory after the repair is just an argument.

Keeping it running after turnover

Some defects are not the crew's fault, they are the owner doing the wrong thing to a sound slab, so the handoff includes the rules. The big one is deicing salt on new exterior concrete. A slab needs to reach maturity before its first freeze-and-deicer exposure, so a new exterior pour should go through its first winter without deicing salts, using sand for traction instead. Salt a green slab and it scales no matter how well it was placed.

The rest of the handoff is short. Seal exterior flatwork and keep it sealed, so water and chlorides stay out of the surface. Keep the joints sealed so water does not get under the slab and feed curling and subgrade problems. Do not let the surface stay saturated, especially going into freeze-thaw. And tell the owner which defects are cosmetic, so a little crazing on a sound slab does not turn into a panicked callback, and which ones, like spreading scaling or a hollow-sounding floor, actually need attention.

Put it in writing at turnover. The owner inherits the slab and the responsibility for how it is treated, and the difference between a floor that lasts and one that scales out is often what got done to it in the first year, not what the crew did the week of the pour.

What to document

Build a record for each defect that names what it is, the cause it points to, whether it is cosmetic or structural, the fix, and the process change that prevents the next one. That record is what turns a callback into a closed loop instead of a repeat. The table below is the quick diagnosis, the same one a foreman carries in their head, laid out so a PM or an owner can read it.

Capture it against the slab with the pour and area, the date, the mix, the weather, and who finished and cured it, then tie each defect to its row below. The cause column is the one that matters most, because it is the column that decides whether the next pour repeats the failure or ends it.

Common mistakes

• Finishing the bleed water into the surface, leaving a weak skin that dusts, scales, blisters, and delaminates.
• Placing exterior concrete with no air entrainment, so it scales the first winter under freeze-thaw and deicers.
• Premature troweling or sealing the top before the concrete is done bleeding and releasing air.
• Hard steel-troweling air-entrained concrete, which traps the air and delaminates or blisters the surface.
• Leaving the slab with no cure, so the surface dries before it hydrates and comes up dusting and crazed.
• Blaming the mix for a finishing defect, or the finisher for a mix defect, and fixing the wrong thing.
• Running a high water-to-cement mix and expecting good finishing to save a soft surface.
• Overlaying a scaled or delaminated surface without removing the failed layer, so the overlay debonds too.
• Letting the owner deice or salt a new exterior slab its first winter before it has reached maturity.

Field checklist

Standards and references

The project specification, the mix supplier, and where the slab is structural the engineer govern, and everything below is the framework those sit on. Where the contract is stricter or more specific, it wins.

ACI 302, the guide for concrete floor and slab construction, is the main reference for the finishing practices that cause or prevent dusting, scaling, blistering, and delamination, including the recommendation against hard-troweling air-entrained concrete. ACI 201, the guide to durable concrete, covers freeze-thaw and deicer scaling, air entrainment, and the durability side of the surface. ACI 305 covers hot-weather concreting and sets the evaporation-rate action level near 0.2 lb per square foot per hour for plastic shrinkage, and ACI 306 covers cold-weather work. ACI 308 governs curing, the stage behind dusting, crazing, and scaling. The exact document numbers and contents shift between editions, so confirm the edition the project adopted before citing a provision.

On the materials and test side, ASTM C309 covers the liquid membrane-forming curing compounds, and ASTM E1155 is the test for FF flatness and FL levelness numbers that a flat floor and its curling are judged against. The NRMCA CIP series and the Portland Cement Association cover the individual defect causes plainly and are the practical references for tracing a defect to its cause. For a suspected alkali-silica reaction behind popouts, a petrographic examination of the aggregate is the diagnostic step. Name the standard that controls the point, verify the edition, and let the project specification and the mix supplier override any rule of thumb.

Units, terms, and conversions

Defect names get mixed up on the slab, and the wrong name sends a crew chasing the wrong cause, so keep them straight. Scaling and spalling are not the same: scaling is the thin top layer flaking off from freeze-thaw and finishing, while spalling is a deeper break, often around reinforcement or an impact. Crazing is shallow cosmetic map cracking, not the wider structural cracks that come from drying shrinkage at the joints.

The few numbers worth carrying: surface defect depths run in inches and millimeters, where 1/8 in is about 3 mm, so crazing under 1/8 in and a scaled layer of 1/8 to 3/16 in are shallow surface failures. The plastic shrinkage action level is 0.2 lb per square foot per hour of evaporation. Water-to-cement ratio is the unitless ratio that drives surface strength, and FF and FL flatness numbers are unitless and higher is flatter.

Scaling

Flaking of the top 1/8 to 3/16 in from freeze-thaw and deicers on under-aired or over-finished concrete

Spalling

A deeper surface break than scaling, often at reinforcement or from impact, not a finishing defect

Crazing

A shallow random map of hairline cracks under about 1/8 in deep; cosmetic, from fast surface drying

Dusting

A soft, powdery surface that never hardened, from bleed water finished in or carbonation

Blister

A small hollow bump from bleed water or air trapped under a prematurely sealed surface

Delamination

A thin top layer debonded over a wide area, hollow-sounding, from troweling over bleed or air

Popout

A conical hole with fractured aggregate at the bottom, from unsound, reactive, or frozen aggregate

Curling

Upward warping of slab edges from the top drying and shrinking faster than the bottom

Plastic shrinkage crack

An early surface crack opened before set when evaporation outruns the bleed rate

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• How long do you cure concrete?
• When do you start finishing concrete?
• What is a concrete admixture?
• What is the difference between a mechanical and an adhesive anchor?
• What is the difference between integral color and stain?